Generative AI and Agentic Operations: Reliable, Scalable Safe Systems
暫譯: 生成式 AI 與代理操作:可靠、可擴展的安全系統

Chilukuri, Krishna K.

  • 出版商: Wiley
  • 出版日期: 2026-09-22
  • 售價: $4,000
  • 貴賓價: 9.5$3,800
  • 語言: 英文
  • 頁數: 160
  • 裝訂: Hardcover - also called cloth, retail trade, or trade
  • ISBN: 1394451733
  • ISBN-13: 9781394451739
  • 相關分類: AI Coding
  • 尚未上市,無法訂購

商品描述

Build, scale, and govern the next generation of autonomous AI agents

The real AI revolution is not in the models themselves, but in their application to the messy, complex, and vital work of running a business. The real challenge is not "how does an agent work?" but "how do I manage a team of people and agent operations?"

Generative AI and Agentic Operations: Reliable, Scalable Safe Systems provides the missing blueprint for this transition. It offers a practical, operator-first guide to building and managing autonomous agents that don't just replace rote tasks, but fundamentally augment and elevate human expertise. Central to this book is the Agentic Ops Lifecycle, a rigorous five-phase framework designed to take you from a fragile prototype to a resilient, production-grade system. Moving beyond traditional MLOps, this guide breaks down the "black box" of LLM reasoning and provides a structured approach to defining, building, evaluating, productionizing, and maintaining a non-deterministic digital workforce.

The book also provides:

  • Tactics for process re-engineering, showing how to consolidate complex workflows and unlock massive operational scale using agentic tools and reasoning loops.
  • Proven strategies for managing a blended "Human+Agent" workforce, shifting human operators from task execution to system supervision and exception handling.
  • A modernized approach to operational metrics and scorecards, moving beyond traditional handle times to measure holistic system quality, cost, and automation rates.
  • Essential frameworks for navigating the ethical and regulatory "grey areas" of AI, including mitigating systemic bias, halting "runaway agents," and building robust security guardrails.
  • Practical insights and real-world case studies drawn from the front lines of transforming complex, large-scale data operations at Google Maps.

Written for the operations leaders, managers, and team leads tasked with integrating AI into their workflows, this book is also essential reading for the product managers and engineers designing these systems. It delivers the strategic vision and the operational depth needed to successfully navigate the Agentic Age.

商品描述(中文翻譯)

**建立、擴展和管理下一代自主 AI 代理**

真正的 AI 革命不在於模型本身,而在於它們在經營企業這一繁雜、複雜且至關重要的工作中的應用。真正的挑戰不是「代理是如何運作的?」而是「我如何管理一個人員團隊和代理操作?」

《生成式 AI 與代理操作:可靠、可擴展的安全系統》提供了這一轉型所需的缺失藍圖。它提供了一本以操作為先的實用指南,幫助建立和管理自主代理,這些代理不僅僅是取代重複性任務,而是根本上增強和提升人類專業知識。本書的核心是代理操作生命週期(Agentic Ops Lifecycle),這是一個嚴謹的五階段框架,旨在將您從脆弱的原型帶到堅韌的生產級系統。超越傳統的 MLOps,本指南解構了大型語言模型(LLM)推理的「黑箱」,並提供了一種結構化的方法來定義、構建、評估、產品化和維護一個非確定性的數位勞動力。

本書還提供:
- **流程再造的戰術**,展示如何整合複雜的工作流程,並利用代理工具和推理循環解鎖巨大的操作規模。
- **管理混合「人+代理」勞動力的經驗策略**,將人類操作員的角色從任務執行轉變為系統監督和例外處理。
- **現代化的操作指標和評分卡方法**,超越傳統的處理時間,衡量整體系統質量、成本和自動化率。
- **導航 AI 的倫理和監管「灰色地帶」的基本框架**,包括減輕系統性偏見、阻止「失控代理」和建立穩健的安全防護措施。
- **實用見解和真實案例研究**,來自於在 Google Maps 轉型複雜、大規模數據操作的前線。

本書是為負責將 AI 整合到工作流程中的運營領導者、經理和團隊負責人而寫的,對於設計這些系統的產品經理和工程師來說,也是必讀之作。它提供了成功導航代理時代所需的戰略願景和操作深度。

作者簡介

Krishna K. Chilukuri is the Senior Manager for Strategy and Operations at Google Maps, where he is responsible for transforming Maps Data Operations using GenAI and Agentic solutions. With over thirteen years at Google and prior experience at Dassault Systems and other multinational companies, he brings deep operational expertise to the intersection of artificial intelligence and business process transformation.

作者簡介(中文翻譯)

Krishna K. Chilukuri 是 Google Maps 的策略與運營高級經理,負責利用 GenAI 和 Agentic 解決方案轉型地圖數據運營。在 Google 工作超過十三年,並曾在達索系統(Dassault Systems)及其他跨國公司任職,他在人工智慧與商業流程轉型的交匯處擁有深厚的運營專業知識。